Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    phrases related to completing or finishing tasks

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to concluding or finalizing actions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ogi
    -0.72
     cited
    -0.60
    oi
    -0.60
     Lite
    -0.59
     risked
    -0.59
     mistaken
    -0.59
    gey
    -0.58
     payable
    -0.57
    ogly
    -0.57
     deficient
    -0.57
    POSITIVE LOGITS
    olars
    0.69
    ths
    0.66
    \\\\\\\\\\\\\\\\
    0.65
     preparations
    0.65
    joy
    0.65
    sett
    0.65
    igham
    0.65
    night
    0.63
     steam
    0.62
    ocrine
    0.62
    Act Density 0.133%

    No Known Activations